# Pearl reports low proof overhead on two production-scale model setups

https://www.learnpearl.org/news/pearl-production-overhead-benchmarks/ · August 6, 2026 · Analysis

Low proof overhead is necessary for useful-mining economics, but it doesn’t measure outside demand or the useful share.

## The reported result

- **Llama 70B · 4×H200:** +5.08% · End-to-end overhead, Pearl-reported
- **DeepSeek V3.2 · 8×H200:** +3.9% · End-to-end overhead, Pearl-reported

Pearl Research Labs says its proof-of-useful-work kernel adds 5.08% end-to-end overhead on a Llama 70B setup using four H200 GPUs, and 3.9% on DeepSeek V3.2 using eight. The current path runs on integer hardware through 7-bit weight-and-activation (W7A7) quantization; a floating-point version is in development.

## Why the measurement matters

If producing proofs ate a large slice of model-serving capacity, block rewards could fail to cover the cost. Single-digit overhead makes the dual-output thesis more plausible for compatible workloads.

## What would make the claim stronger

- Reproducible benchmark code and workload configuration.
- Independent replication across more models and GPU generations.
- Latency, throughput, power and dollar cost reported together.
- Production usage and savings after rewards are sold at realistic prices.

> **Whose numbers these are.** These are Pearl’s measurements, not an independent industry benchmark.

## Sources

- [Proof of Useful Work from the Ground Up](https://pearlresearch.ai/research/blog/proof-of-useful-work) (Pearl Research Labs, primary)